An investigation was undertaken regarding the determination of desirable levels of statistical power for implementation in dental caries clinical trials. The chief concern was to determine sufficiently large cohort sizes so as to be able to reliably detect specified differences in therapeutic efficacy levels among prophalactic agents for dental caries. Another concern was to investigate the consequences of utilizing an insufficient number of subjects in such trials. Statistical tables were also derived for use in the determination of such cohort sizes in terms of the error probability magnitudes and the anticipated efficacy levels.